INB-Lunch-Seminar

Matrix and Tensor Factorizations in Vision

Christian Bauckhage

This talk will discuss the use of advanced (multi)linear algebra in computer vision. Our group works on problems in the areas of object recognition, activity recognition, and categorization of the aesthetic content of visual data. We found that matrix and tensor factorization methods offer great potential in this regard. In the talk, we shall provide a gentle introduction to the techniques and present application examples from our recent work.